Potent N-(1,3-Thiazol-2-yl)pyridin-2-amine vascular endothelial growth factor receptor tyrosine kinase inhibitors with excellent pharmacokinetics and low affinity for the hERG ion channel

Article abstract of DOI:10.1021/jm049697f

A series of N-(1,3-thiazol-2-yl)pyridin-2-amine KDR kinase inhibitors have been developed that possess optimal properties. Compounds have been discovered that exhibit excellent in vivo potency. The particular challenges of overcoming hERG binding activity and QTc increases in vivo in addition to achieving good pharmacokinetics have been acomplished by discovering a unique class of amine substituents. These compounds have a favorable kinase selectivity profile that can be accentuated with appropriate substitution.

Full text of DOI:10.1021/jm049697f

J. Med. Chem. 2004, 47, 6363-6372
6363
Potent N-(1,3-Thiazol-2-yl)pyridin-2-amine Vascular Endothelial Growth Factor
Receptor Tyrosine Kinase Inhibitors with Excellent Pharmacokinetics and Low
Affinity for the hERG Ion Channel
Mark T. Bilodeau,*,† Adrienne E. Balitza,† Timothy J. Koester,† Peter J. Manley,† Leonard D. Rodman,†
Carolyn Buser-Doepner,‡ Kathleen E. Coll,‡ Christine Fernandes,‡ Jackson B. Gibbs,‡ David C. Heimbrook,‡
William R. Huckle,‡ Nancy Kohl,‡ Joseph J. Lynch,| Xianzhi Mao,‡ Rosemary C. McFall,‡ Debra McLoughlin,§
Cynthia M. Miller-Stein,§ Keith W. Rickert,‡ Laura Sepp-Lorenzino,‡ Jennifer M. Shipman,‡ Raju Subramanian,§
Kenneth A. Thomas,‡ Bradley K. Wong,§ Sean Yu,§ and George D. Hartman†
Departments of Medicinal Chemistry, Cancer Research, Drug Metabolism and Pharmacology, Merck Research Laboratories,
P.O. Box 4, West Point, Pennsylvania 19486
Received April 23, 2004
A series of N-(1,3-thiazol-2-yl)pyridin-2-amine KDR kinase inhibitors have been developed that
possess optimal properties. Compounds have been discovered that exhibit excellent in vivo
potency. The particular challenges of overcoming hERG binding activity and QTc increases in
vivo in addition to achieving good pharmacokinetics have been acomplished by discovering a
unique class of amine substituents. These compounds have a favorable kinase selectivity profile
that can be accentuated with appropriate substitution.
Introduction
Vascular endothelial growth factor (VEGF) is a selec-
tive mediator of endothelial cell mitogenesis and chemo-
taxis that has been identified as a principal factor in
the induction of angiogenesis.1 It has been implicated
in a variety of pathological conditions that are angio-
genesis-dependent including cancer, arthritis, diabetic
retinopathy, and some inflammatory diseases. VEGF
Figure 1. Lead structures.
has been a particular focus of antiangiogenic therapies
strategy that was successful in other lead series, we
against malignant disease. Its expression has been
sought to increase cell potency by the incorporation of
shown to be up-regulated in a variety of tumors and has
a solubility-enhancing amine substituent. A benzylic
been reported to be a prognostic indicator of tumor
pyrrolidine was thus incorporated at the 5-pyridyl
progression and/or decreased survival.2 VEGF activation
position to produce 2 (Figure 1) with excellent potency
of the transmembrane receptor tyrosine kinase-linked
in the cell-based assay. Compound 2, however, had poor
KDR (VEGFR-2) induces vascular endothelial cell mi-
pharmacokinetics and was a potent binder to the
tosis and migration.3 Therapeutic approaches that
potassium channel encoded by the human ether-a-go-
target VEGF or KDR have shown positive results in
go-related gene (hERG) which is an inward rectifying
preclinical tumor models and have been actively under-
potassium channel involved in ventricular repolariza-
going clinical evaluation. Bevacizumab, a VEGF anti-
tion. This channel underlies the rapid component of the
body was reported to be efficacious in Phase III clinical
delayed rectifier K+ current (IKr), and blockade of IKr
trials.4 Clinical evaluations have also been undertaken
may lead to prolongation of the QT interval and to
with an anti-KDR antibody, a soluble VEGF decoy-
potentially fatal ventricular arrhythmias including tor-
receptor, as well as small molecule inhibitors of KDR
sades de pointes.8 Many marketed drugs are known to
kinase activity.5
block IKr.9 In several cases this has led to regulatory
We have previously disclosed N-(1,3-thiazol-2-yl)-
pyridin-2-amines as a potent lead class of KDR kinase
inhibitors.6 In this paper, we present the development
of this series toward compounds with ideal drug char-
acteristics. The initial lead compound 1 (Figure 1), while
potent in the KDR enzyme assay, had poor potency in
an endothelial cell mitogenesis assay.7 Following a
intervention and withdrawal of drugs from the market.
This paper will address the optimization to decrease
affinity for the hERG channel and to achieve compounds
with excellent in vivo potency and excellent pharmaco-
kinetics in three species.
Chemistry. (2-Chloropyridin-4-yl)methanol (3, Scheme
1)10 was treated with tert-butyldimethylsilyl chloride
and imidazole to provide 4. Compound 4 was aminated
with benzophenone imine, and the crude material was
hydrolyzed with hydroxylamine to provide 5.11 The
aminopyridine was then reacted with NaH and a
2-chlorothiazole (R2 ) CN or Ph) to provide the N-(1,3-
thiazol-2-yl)pyridin-2-amine 6. The resulting coupled
products 6 (R1 ) H) were desilylated with HF-pyridine
* To whom correspondence should be addressed at the Department
of Medicinal Chemistry, Merck Research Laboratories, WP14-2, P.O.
Box 4, West Point, PA 19486. Tel (215) 652-5304; fax (215) 652-7310;
e-mail mark_bilodeau@merck.com.
† Department of Medicinal Chemistry.
‡ Department of Cancer Research.
§
Department of Drug Metabolism.
| Department of Pharmacology.
10.1021/jm049697f CCC: $27.50 © 2004 American Chemical Society
Published on Web 10/29/2004

6364 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Bilodeau et al.
Scheme 1
Table 1.
a For determinations where n > 2, the standard deviation is given. For determinations where n ) 2 the number is shown in parentheses
and in these cases the determinations were within ( 25% of the mean. The remainder of the data is from single determinations. b Average
of two dogs after iv dosing. c Compounds 12 and 13 were produced as part of a HPLC-purified library and were characterized by exact
mass measurement.
verted according to the method employed for 4 (R1 )
H) to ultimately provide the products 8 (R1 ) Me).
Scheme 2
Results and Discussion
The lead compound 2 was potent against KDR (Table
1) as well as potent in two cell assay formats. The two
cell-based formats7 are an endothelial cell mitogenesis
assay (Endo. Mit.) measuring inhibition of VEGF-
induced DNA synthesis and a cell-based KDR autophos-
phorylation inhibition assay (KDR Phos.) in HEK293
cells stably transfected to express high levels of full-
length human KDR. However, 2 was found to be very
potent in an hERG binding assay and exhibited poor
pharmacokinetics in dog with a clearance well above
hepatic blood flow (Table 1). The lipophilicity of 2 was
high (log P ) 3.59) and was considered to be a contrib-
uting factor to its hERG binding and poor pharmaco-
kinetics. By employing the pyridine substitution pattern
outlined in Scheme 1, we initially sought to explore
whether variation of the amine substituent would
favorably impact the properties of the series.
and converted to the chloromethylpyridines 7 (R1 ) H)
with POCl3 and DMF. Reaction of the chloride with a
variety of amines at ambient temperature in DMSO
resulted in the final products 8.
The synthesis of a key intermediate for the case where
R1 ) Me is shown in Scheme 2. Methyl 3-methylisoni-
cotinate (9) was oxidized with MeReO3 and hydrogen
peroxide,12 and the resulting pyridine N-oxide was
treated with POCl3 to afford methyl 2-chloro-3-methyl-
isonicotinate 10 (produced as a 2:1 ratio with the
regioisomeric methyl 2-chloro-5-methylisonicotinate).
Reduction of the ester with LiBH4 and silylation as
above resulted in 11. Intermediate 11 was then con-
The first clue that the undesirable properties of 2
could be altered by choice of pendant amine came from
a small library of analogues. This library produced 40
compounds related to 12 (Table 1) where the amine
substituent was varied. All changes of this 4-(amino-

VEGF Receptor Tyrosine Kinase Inhibitors
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25 6365
Table 2.
a For determinations where n > 2, the standard deviation is given. For determinations where n ) 2, the number is shown in parentheses
and the individual determinations are within (25% of the mean unless noted otherwise. The remainder of the data is from single
determinations. b The two determinations are within (50% of the mean
methyl)pyridine moiety did not effect potency or kinase
selectivity which is consistent with our postulated
binding mode of these compounds where the amine
extends out of the ATP-binding cleft and is exposed to
solvent.6 However, this library provided an insight into
modifying both hERG binding activity and metabolism.
Compound 12 is representative of all but one amine in
the library, with excellent enzyme and cell-based po-
tency but, like 2 displayed very potent hERG binding
(IP ) 22 nM). This compound was also metabolized very
rapidly in a rat liver microsomal incubation (t1/2 ) 8
min). The one exception to these observations that was
found in this library was 13 (Table 1). This compound,
bearing an N-acetylpiperazine moiety displayed signifi-
cantly lower, albeit still potent, hERG binding (IP ) 240
nM) and a greater resistance to metabolism (t1/2 ) 251
min). A subsequent analogue of 13 bearing a methane-
sulfonamide moiety (14, Table 1) showed a similar
potency profile to 13 and was additionally found to have
a moderate clearance in dog, indicating the significant
improvement these unique substituted piperazine end-
groups provide relative to 2.
This discovery was extended to produce compounds
with excellent pharmacokinetics and very low affinity
for hERG. We focused on changes to the lipophilic
thiazoyl-5-phenyl moiety which was postulated to be a
key contributor to both the hERG binding and the poor
pharmacokinetics. In initial SAR studies it was found
that the phenyl moiety could be replaced with a cyano
moiety with only a moderate loss of potency.6 When a
nitrile was employed to replace the phenyl in 13, a
potent inhibitor was produced (Table 2, 15). This
compound had excellent enzyme potency and cell-based
activity in the two assay formats. Moreover, this mol-
ecule displayed significantly lower affinity for hERG and
good physical properties, with a low log P of 2.06 and a
significant free fraction in human plasma (89.4% plasma
bound). Compound 15 also showed excellent dog phar-
macokinetics (Table 3), with a long half-life, low clear-
ance and an excellent oral bioavailability. This optimal
profile is dependent on the acetamide as either an
unsubstituted piperazine (16) or a morpholine (17)
produced compounds that were still potent but had
slightly greater affinity for hERG and significantly
poorer dog pharmacokinetics (Cl ) 24.2 and 12.4 mL/

6366 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Table 3. Dog Pharmacokinetic Data for Selected Compoundsa
Bilodeau et al.
Table 4. Pharmacokinetic Data for 20a
compd
t1/2 (h)
Cl (mL/min/kg)
Vdss (L/kg)
F (%)
PK parameter
dogb
ratc
Rhesusd
15
16
17
18
19
20
21
22
23
7.2
7.9b
2.5
1.2
24.2b
12.4
2.6
0.72
21.9b
1.0
0.34
0.88
0.72
0.47
0.96
1.2
87
100
nd
nd
nd
99
t1/2 (h)
17.4
0.51
0.72
99
1.9 ( 0.35
13.3 ( 1.0
1.5 ( 0.19
49
1.7
17.2
1.3
Cl (mL/min/kg)
Vdss (L/kg)
F (%)
2.9
43
2.0
17.4
6.6
6.3
7.0
7.4
a Dosed iv as a DMSO solution and po as a solution in 0.05 M
citric acid (aq). b Average of two dogs dosed at each of 1 mg/kg iv
and 1 mg/kg po. The individual determinations are within (25%
of the mean. c Average of three rats (with standard deviation)
dosed at each of 2 mg/kg iv and 10 mg/kg po. d Average of two
rhesus dosed at each of 1 mg/kg iv and 1 mg/kg po. The individual
determinations are within (25% of the mean.
0.51
0.88
1.8
61
95
75
2.0
a Average of two dogs dosed iv (0.25 mg/kg for 17, 18, and 19
or 1 mg/kg for all others) as a solution in DMSO and po (1 mpk)
as a solution in acidified methocel or 0.05 M citric acid (aq). The
individual determinations are within (25% of the mean unless
noted otherwise. b The two determinations are within (50% of the
mean.
Table 5. Pharmacokinetic Data for 23a
PK parameter
dogb
ratc
Rhesusd
t1/2 (h)
7.0
2.0
1.2
75
2.4 ( 0.72
3.0 ( 0.45
0.6 ( 0.06
8.8
1.7
1.3
92
Cl (mL/min/kg)
Vdss (L/kg)
F (%)
min/kg, respectively, Table 3). The acetamide however
could be replaced by other polar carbonyl-containing
moieties to maintain or enhance the profile. The termi-
nal urea 18 was potent and had very low affinity for
hERG with still good dog pharmacokinetics albeit with
a diminished t1/2 (2.9 h). The N,N-dimethylurea ana-
logue 19 had significantly increased affinity for hERG
and an even higher clearance. Interestingly, a metabo-
lite of 19, the monomethyl urea 20, displayed a lower
affinity than 19 for hERG and resulted in exceptional
dog pharmacokinetics, with a t1/2 of 17.4 h, very low
clearance of 0.51 mL/min/kg, and outstanding bioavail-
ability (99%). Other polar piperazine amides proved to
have excellent properties, such as the hydroxyacetamide
21, and the optimal properties were not limited to
piperazine derivatives. The piperidine 22, with a polar
methylsulfonyl moiety also displayed low hERG affinity
and excellent pharmacokinetic behavior in dogs.
In one example of additional substitution of the
pyridine ring, we appended a methyl group to the
3-position of 20, resulting in 23. This compound was
very similar in potency to 20, with a slightly greater
affinity for hERG. It also had excellent pharmacokinet-
ics in dogs (Table 3) albeit with a slightly increased
clearance and lower t1/2. The introduction of this methyl
group lead to a different profile than 20 with respect to
both pharmacokinetics in other species and to kinase
selectivity (vide infra).
73
a Dosed iv as a DMSO solution and po as a solution in 0.05 M
citric acid (aq). b Average of two dogs dosed at each of 1 mg/kg iv
and 1 mg/kg po. c Average of three rats dosed at each of 2 mg/kg
iv and 10 mg/kg po. d Average of two rhesus dosed at each of 1
mg/kg iv and 1 mg/kg po.
erate clearances and half-lives close to 2 h were observed
(Table 4). After oral administration, 20 was well ab-
sorbed in both rat and rhesus. Compound 20 was not a
potent reversible inhibitor of the major human CYP
isozymes; IC50 values for CYP 2C9, 2D6 and 3A4 were
all greater than 50 µM.
Compound 23, the 3-methylpyridine analogue of 20,
had slightly diminished but still excellent pharmacoki-
netic behavior in dog, with a long half-life of 7.0 h and
low clearance (2.0 mL/min/kg) (Table 5). Interestingly
the introduction of the 3-methyl group significantly
enhanced pharmacokinetics in the rat and rhesus rela-
tive to 20. Following an iv dose, the mean plasma
clearance was low in both the rat and rhesus (3.0 and
1.7 mL/min/kg, respectively), and the t1/2 was slightly
improved in the rat (2.4 h) and significantly improved
in the rhesus (8.8 h). High bioavailabilities were ob-
served in both rat and rhesus relative to 20. It is also
not a potent inhibitor of the major CYP isozymes; IC50
values for CYP 2C9, 2D6, and 3A4 were all greater than
100 µM.
While hERG binding is an imperfect predictor for in
vivo QT prolongation, it is one of the few available high
throughput assay techniques.13 As a critical follow-up
to binding measurements, several of these compounds
were evaluated for their ability to prolong QTc in vivo
in anesthetized dogs. While 2 caused a 10% increase in
the QTc interval at a plasma concentration of 0.59 µM,
the optimized compounds with greatly diminished hERG
binding showed a significantly reduced effect in vivo.
For 15 a 10% increase in QTc was not observed when
dosed up to 60 µM plasma levels. Additionally, com-
pound 23 did not show a 10% increase in QTc up to 22
µM plasma levels. Thus, these optimized compounds are
expected to present little risk of QT prolongation in vivo.
The pharmacokinetic behavior of 20 was profiled in
three species (Table 4). As mentioned above, the dog
pharmacokinetics for 20 were excellent, with a long half-
life, low clearance, and high oral bioavailability. Some-
what poorer pharmacokinetics were observed for 20 in
both rat and rhesus. Following iv administration mod-
To profile the in vivo metabolism and elimination of
20, a 14C-labeled compound was dosed iv to rats, dogs,
and rhesus. Biotransformation was found to be the
major route of elimination in all of these species; the
fraction of the dose excreted unchanged in bile and urine
was <16%. N-Glucuronidation of the parent compound
was identified as the major route of metabolism in both
rat and rhesus (37.6 and 63.3% of iv dose, respectively)
but was a minor component in dogs. Mass spectrometric
and NMR analysis has shown the structure of the
N-glucuronide metabolite to be 24 (Figure 2), where
glucuronidation has occurred on the central nitrogen.
Other minor metabolites in rats and monkeys were
products of oxidation and N-demethylation. In dogs, 20
was eliminated in the form of several oxidative metabo-
lites, of which N-demethylated products were the most
abundant.
Consistent with the in vivo observations, studies
conducted with hepatocytes also showed species-de-
pendent metabolism of [14C]20. N-Glucuronidation was

VEGF Receptor Tyrosine Kinase Inhibitors
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25 6367
Figure 2. N-Glucuronide of [14C]20.
Table 6. Kinase Selectivity Expressed as the Ratio between
Figure 3. Determination of in vivo IC50 for the inhibition of
mouse lung KDR phosphorylation by 20.
the IC50 for the Given Kinase and the IC50 for KDR
kinase
20
23
Flk-1
Flt-1
4
25
28
3
81
83
receptor tyrosine kinases, the c-Src nonreceptor tyrosine
kinase, and one of the cyclin-dependent kinases, CDK4.
A very high degree of selectivity is displayed against
the EGFR, IGF-1R, and insulin receptor tyrosine ki-
nases, other Src family nonreceptor tyrosine kinases,
Lyn and FynT, and the other cyclin-dependent kinase
tested (CDK2).
Flt-3
Flt-4
PDGFRâ
c-kit
c-fms
FGFR1
FGFR2
EGFR
insulinR
Src
Lyn
Lck
FynT
CDK4
CDK2
4.2
3.5
18
45
84
104
41
14
40
81
581
271
In our initial SAR studies we observed that substitu-
tion of the pyridine ring can enhance selectivity.6 The
most consistent observation of this effect has occurred
with the introduction of the 3-methyl substituent on the
pyridine ring. When such a substituent was introduced
into 20, resulting in 23, an enhanced level of selectivity
was observed. Compound 23 maintained comparable
selectivity to 20 versus Flk-1, Flt-4, PDGFRâ, and c-kit.
However, moderate improvements in selectivity were
seen against Flt-1, Flt-3, and the cyclin dependent
kinases tested. Most notably, selectivity was signifi-
cantly enhanced against FGFR-1, FGFR-2, and Src.
In Vivo Activity. A whole animal pharmacodynamic
assay was used to evaluate the in vivo activity of 20.7
The assay measures the inhibition of VEGF-induced
tyrosine phosphorylation of mouse lung KDR as a
function of compound dose and time after dosing. This
method produces an in vivo IC50 value for a KDR
inhibitor as a function of compound plasma concentra-
tion. A curve was generated comparing phosphorylated
KDR levels in the lung between compound treated and
untreated lungs after VEGF-stimulation by tail vein
injection of human VEGF165. Compound 20 inhibits
VEGF-stimulated mouse lung KDR autophosphoryla-
tion in a dose-dependent manner with an IC50 value of
90 ( 20 nM as shown in Figure 3. This value is
indicative of a high level of in vivo activity of the
compound, particularly when taking into consideration
the 4-fold shift in activity that was observed between
KDR kinase activity and Flk-1 (the mouse isoform of
KDR, Table 6).
>1600
560
84
>1600
734
474
869
nd
950
850
61
2670
3728
185
328
1140
the major pathway observed in rat and monkey hepa-
tocyte incubations, with oxidation and N-demethylation
also observed in these species. However, in dog and
human hepatocytes, N-demethylation and oxidation
appeared to be the dominant metabolic routes while the
N-glucuronide conjugate of 20 was not detected in dog
and human hepatocyte preparations.
The rate of metabolism in hepatocytes followed the
of the rank order monkey > rat . dog. This agreed well
with the rank order of in vivo plasma clearance: mon-
keys (17 mL/min/kg) > rats (13 mL/min/kg) . dog (0.51
mL/min/kg). The rate of 20 metabolism in human
hepatocytes from three different donors was similar to
that in dogs. In addition, plasma protein binding is
similar across all of these species. It is expected that
20 would be a low clearance compound in humans as it
is in dogs.
Given the observation of significant glucuronidation
of 20 in rat and rhesus, it is interesting to note that 23,
with the 3-methylpyridine substituent, has significantly
improved pharmacokinetics in these species relative to
20. It is likely that the 3-methylpyridine substitution
sterically blocks the glucuronidation event and leads to
the observed improvement in pharmacokinetics.
Kinase Selectivity. The kinase selectivity profiles
for 20 and 23 are shown in Table 6. The values in the
table are expressed as a ratio of the given kinase IC50
to the KDR kinase IC50. Compound 20 shows a slight
degree of selectivity against Flt-4 (VEGFR-3), a receptor
homologue primarily expressed on lymphatic endothe-
lial cells, and Flk-1 (the mouse isoform of KDR), both
of which are highly homologous to human KDR. It was
found to have moderate selectivity against the highly
homologous c-kit, Flt-1, Flt-3, c-fms, and PDGFRâ
enzymes. It also shows moderate to good selectivity
against the less homologous FGFR-1 and FGFR-2
Conclusion
A series of N-(1,3-thiazol-2-yl)pyridin-2-amines has
been optimized to provide KDR kinase inhibitors with
optimal properties. Initial problems encountered with
cell-potent compounds in this series involved poor
pharmacokinetics and potent hERG binding activity.
The problems associated with hERG binding and phar-
macokinetics were resolved by the variation of two key
moieties in the molecules. The discovery of a unique
series of amine substituents, coupled with the replace-
ment of a lipophilic phenyl with a cyano substituent,
lead directly to the solution of these challenges. The

6368 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Bilodeau et al.
resulting solution was taken to neutral pH with 1 M HCl (aq)
and filtered. The residue was purified by flash column chro-
matography using 20% EtOAc in hexane to provide 1.31 g
(79%) of the pure title compound. 1H NMR (CDCl3): δ 9.09
(bs, 1H), 8.32 (d, 1H, J ) 5.2 Hz), 7.62 (s, 1H), 7.56 (d, 2H, J
) 7.4 Hz), 7.38 (t, 2H, J ) 7.6 Hz), 7.26 (overlapping with
CHCl3, 1H), 6.90 (s, 1H), 6.82 (d, 1H, J ) 5.2 Hz), 4.75 (s, 2H),
0.96 (s, 9H), 0.14 (s, 6H). mp 207 °C.
most highly optimized compound 20 was shown to have
low affinity for hERG, excellent pharmacokinetics, and
good activity in an in vivo assay of KDR kinase inhibi-
tion. Detailed pharmacokinetic analysis suggested that
20 is likely to have a low clearance in humans.
Experimental Section
[2-(5-Phenylthiazol-2-ylamino)pyridin-4-yl]metha-
nol. [4-(tert-Butyldimethylsilanyloxymethyl)pyridin-2-yl](5-
phenylthiazol-2-yl)amine (0.805 g, 2.03 mmol) was dissolved
in 10 mL of THF, and the resulting solution was cooled to 0°C.
Hydrogen fluoride-pyridine (Aldrich, HF ∼70%, pyridine
∼30%) 1.20 mL was added. After 1 h the reaction was allowed
to warm to RT. The THF was removed in vacuo, and the
residue was diluted with sat Na2CO3 (aq). The resulting
precipitate was filtered to provide the pure title compound as
a white solid (0.560 g, 98% yield). 1H NMR (DMSO-d6): δ 11.35
(bs, 1H), 8.25 (d, 1H, J ) 5.2 Hz), 7.79 (s, 1H), 7.59 (d, 2H, J
) 7.4 Hz), 7.39 (t, 2H, J ) 7.6 Hz), 7.25 (t, 1H, J ) 7.3 Hz),
7.08 (s, 1H), 6.86 (d, 1H, J ) 5.2 Hz), 5.42 (bs, 1H), 4.51 (s,
2H). Mp 236-237 °C.
(4-Chloromethylpyridin-2-yl)(5-phenylthiazol-2-yl)-
amine (7). [2-(5-Phenylthiazol-2-ylamino)pyridin-4-yl]metha-
nol (0.500 g, 1.77 mmol) was stirred in anhydrous CH2Cl2 (5
mL) under N2. N,N-Dimethylformamide (0.137 mL, 1.76 mmol)
was added followed by the addition of phosphorus oxychloride
(0.165 mL, 1.76 mmol). After 1.5 h the reaction was concen-
trated and quenched by the addition of saturated NaHCO3
(aq). A precipitate formed which was filtered and washed with
water to provide the title compound as a tan solid (0.482 g,
90% yield). 1H NMR (DMSO-d6) δ 11.49 (bs, 1H), 8.34 (d, 1H,
J ) 5.2 Hz), 7.81 (s, 1H), 7.60 (d, 2H, J ) 7.7 Hz), 7.39 (t, 2H,
J ) 7.6 Hz), 7.26 (t, 1H, J ) 7.0 Hz), 7.13 (s, 1H), 6.99 (d, 1H,
J ) 5.3 Hz), 4.77 (s, 2H).
(5-Phenylthiazol-2-yl)(4-pyrrolidin-1-ylmethylpyridin-
2-yl)amine (12). Prepared in a library format: Pyrrolidine
(0.011 g, 0.15 mmol) was dissolved in 0.1 mL of DMSO. (4-
Chloromethylpyridin-2-yl)(5-phenyl-thiazol-2-yl)amine 7 (0.0095
g, 0.031 mmol) dissolved in 0.2 mL of DMSO was added to
the pyrrolidine solution. After 16 h the reaction was directly
purified by reverse phase preparative chromatography. The
resulting product as a TFA salt was characterized by HRMS
[M + H]+ calcd: 337.1487, found: 337.1511. HPLC (method
A): 98.0% purity.
1-{4-[2-(5-Phenylthiazol-2-ylamino)pyridin-4-ylmeth-
yl]piperazin-1-yl}ethanone (13). Prepared in a library
format according to preparation of 12. HRMS [M + H]+ calcd:
394.1702, found: 394.1697. HPLC (method A): 98.5% purity.
4-{[4-(Methylsulfonyl)piperazin-1-yl]methyl}-N-(5-phen-
yl-1,3-thiazol-2-yl)pyridin-2-amine (14). (4-Chlorometh-
ylpyridin-2-yl)(5-phenylthiazol-2-yl)amine 7 (0.053 g, 0.176
mmol) was dissolved in 1 mL of DMSO. Et3N (0.122 mL, 0.88
mmol) was added followed by addition of 1-(methylsulfonyl)
hydrochloride (0.042 g, 0.21 mmol), and the reaction was
stirred at RT. After 30 min the reaction was warmed to 45
°C. After 2 h the reaction was diluted with water, and the
resulting solid was filtered and washed with water. The solid
was purified by reverse phase chromatography to provide 27
mg (28% yield) of the title compound as a trifluoroacetate salt.
1H NMR (DMSO-d6) δ 11.49 (bs, 1H), 8.51 (d, 1H, J ) 5.5 Hz),
7.79 (s, 1H), 7.60 (d, 2H, J ) 7.8 Hz), 7.44 (t, 2H, J ) 7.3 Hz),
7.37 (m, 1H), 7.32 (bs, 1H), 7.25 (d, 1H, J ) 5.5 Hz), 4.31 (s,
2H), 3.50 (bs, 4H), 3.32 (bs, 4H), 2.95 (s, 3H). mp ) 183-184
°C. HRMS [M + H]+ calcd: 430.1731, found: 430.1722. HPLC
(method A): 98.5% purity. HPLC (method B): 98.9% purity.
2-Chlorothiazole-5-carbonitrile. A flame-dried round-
bottom flask under N2 was charged with 150 mL of anhydrous
MeCN. CuCl2 (12.9 g, 95.9 mmol, 1.2 equiv) was added, and
the reaction was maintained in a room-temperature bath. tert-
Butyl nitrite (14.3 mL, 120 mmol, 1.5 equiv) was added
gradually over 10 min. After 10 min, 2-aminothiazole-5-
carbonitrile17 (10.0 g, 79.9 mmol) was added as a solid
gradually. The reaction was stirred at room temp for 4h. The
General Methods. General Experimental Information.
Proton and carbon NMR spectra were recorded on Varian
Unity 400 or VRX-400 spectrometers (400 MHz) or Varian
Unity or Varian Plus (500 MHz) spectrometers. Chemical
shifts are reported in parts per million (δ) downfield from
tetramethylsilane as an internal standard. High-resolution
mass spectra were recorded on a Bruker 3T Fourier transform
ion cyclotron resonance mass spectrometer equipped with
electrospray ionization. The MH+ signal of the analyte of
interest is mass measured against the internal calibrant,
polypropylene glycol (average mass (425). Solvents and re-
agents were obtained from commercial sources and used
without further purification. The reported yields are the actual
isolated yields of purified material and are not optimized.
Flash column chromatography was performed on an Isco
Combiflash system employing prepacked silica columns (Isco
Redi-Sep columns or Applied Separations Speed cartridges).
Preparative reverse phase HPLC was performed on an auto-
mated Gilson system using a gradient of 5% MeCN to 95%
MeCN (0.1% TFA) over 20 min on a Supelcosil ABZ-Plus
column (10 cm × 21.2 mm, 5 µM). Analytical HPLC method
A: Thermoseparation Products HPLC with a gradient from
10% MeCN to 100% MeCN (0.1% H3PO4) in 3 min using a
YMC-Pack ProC18 4.6 × 50 mm column. Analytical HPLC
method B: Waters’ LC/MS comprised of a 2695 LC with a
gradient from 7% MeCN to 100% MeCN (0.05% TFA) in 3 min,
using a YMC PRO LC 3 × 50 mm column. UV analysis was
performed at 215 nM for both methods.
4-({[tert-Butyl(dimethyl)silyl]oxy}methyl)-2-chloropy-
ridine (4). To a solution of (2-chloropyridin-4-yl)methanol10
(16.83 g, 117 mmol) in dry THF (300 mL) were added imidazole
(9.58 g, 140.7 mmol) and TBDMSCl (21.2 g, 140.7 mmol) at
RT. After 6 h imidazole (2.39 g, 35.1 mmol) and TBDMSCl
(5.3 g, 35.1 mmol) were added and stirring continued. After
18 h the mixture was filtered and concentrated. Flash column
chromatography (7% EtOAc/hexanes) gave the title compound
(29.35 g, 97%) as a clear oil: 1H NMR (500 MHz, CDCl3) δ
8.31 (d, J ) 5.13 Hz, 1 H), 7.31 (s, 1 H), 7.15 (m, 1 H), 4.73 (s,
2 H), 0.96 (s, 9 H), 0.12 (s, 6 H).
4-({[tert-Butyl(dimethyl)silyl]oxy}methyl)pyridin-2-
amine (5, R1 ) H). To a solution of 4-({[tert-butyl(dimethyl)-
silyl]oxy}methyl)-2-chloropyridine (29.35 g, 113.8 mmol) in dry
toluene (250 mL) were added NaOtBu (15.3 g, 159.4 mmol),
racemic-BINAP (2.13 g, 3.41 mmol), Pd2(dba)3 (1.04 g, 1.14
mmol), and benzophenone imine (23 mL, 136.6 mmol). The
mixture was degassed (3 × pump/N2) then heated to 80 °C.
After 5 h the mixture was cooled to RT, diluted with Et2O (1
L), and filtered through a pad of Celite. The filtrate was
concentrated. The residue was taken up in 250 mL of MeOH,
and hydroxylamine (15 mL of 50% solution in H2O) was added
at RT. After 18 h the mixture was filtered through a pad of
Celite and concentrated. The residue was taken up in Et2O,
cooled to 0 °C, and filtered. The filtrate was concentrated.
Flash column (gradient, 50-100% EtOAc/hexanes) gave
4-({[tert-butyl(dimethyl)silyl]oxy}methyl)pyridin-2-amine (17.89
g, 66%) as a light yellow solid: 1H NMR (500 MHz, CDCl3) δ
7.99 (d, J ) 5.13 Hz, 1 H), 6.57 (m, 1 H), 6.51(d, J ) 0.74 Hz,
1 H), 4.64 (s, 2 H), 4.39 (bs, 2 H), 0.95 (s, 9 H), 0.11 (s, 6 H).
[4-(tert-Butyldimethylsilanyloxymethyl)pyridin-2-yl]-
(5-phenylthiazol-2-yl)amine (6, R1 ) H, R2 ) Ph). 4-(tert-
Butyldimethylsilanyloxymethyl)pyridin-2-ylamine (1.00 g, 4.19
mmol) was dissolved in 20 mL of anhydrous THF at room
temp, and NaH (60% dispersion, 0.670 g, 16.8 mmol) was
added. When the bubbling stopped, 2-chloro-5-phenylthiazole16
0.739 g (3.78 mmol) was added, and the reaction was heated
to reflux. After 2 h the THF was removed in vacuo, and the

VEGF Receptor Tyrosine Kinase Inhibitors
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25 6369
reaction was poured into 400 mL of 0.5 M HCl (aq). The
mixture was extracted 3× with EtOAc. The organic phases
were dried over Na2SO4, filtered and concentrated to afford
and the resulting precipitate was filtered and washed with
water and dried. This provided 0.130 g (62% yield) of the pure
1
title compound. H NMR (DMSO-d6) δ 12.58 (bs, 1H), 10.18
1
pure desired product (8.40 g, 73% yield). H NMR (CDCl3) δ
(bs, 1H), 8.34 (d, 1H, J ) 5.1 Hz), 8.27 (s, 1H), 7.11 (s, 1H),
7.04 (d, 1H, J ) 5.1 Hz), 3.57 (s, 2H), 3.32 (m, 4H), 3.01 (m,
4H). HRMS [M + H]+ calcd: 301.1235, found: 301.1197. Anal.
(C14H16N6S) C, H, N.
8.04 (s).
2-[4-(tert-Butyldimethylsilanyloxymethyl)pyridin-2-
ylamino]thiazole-5-carbonitrile (6, R1 ) H, R2 ) CN).
4-(tert-Butyldimethylsilanyloxymethyl)pyridin-2-ylamine (16.1
g, 67.5 mmol) was dissolved in 200 mL of anhydrous THF
under N2. NaH (60% suspension, 8.09 g, 202 mmol, 3 equiv)
was added (vigorous bubbling occurs), and the resulting
mixture was stirred for 15 min. 2-Chlorothiazole-5-carbonitrile
(10.1 g, 81.0 mmol) was added, and the reaction was heated
to reflux. After 2 h the reaction was cooled and was quenched
by the addition of water. The THF was removed in vacuo, and
the resulting aqueous solution was adjusted to pH ) 7 by the
addition of 1 M HCl (aq). The resulting precipitate was filtered
and washed with water to provide the desired product in
moderate purity (23.4 g, 101% yield). 1H NMR (CDCl3) δ 10.32
(bs, 1H), 8.33 (d, 1H, J ) 5.3 Hz), 7.99 (s, 1H), 6.96 (s, 1H),
6.91 (d, 1H, J ) 5.3 Hz), 4.78 (s, 2H), 0.98 (s, 9H), 0.16 (s,
6H).
2-(4-Morpholin-4-ylmethylpyridin-2-ylamino)thiazole-
5-carbonitrile (17). The procedure was followed as for 15 but
employing morpholine (0.087 g, 1.00 mmol), 0.8 mL of DMSO,
and 2-(4-chloromethylpyridin-2-ylamino)thiazole-5-carbonitrile
(0.050 g, 0.20 mmol). After preparative reverse phase HPLC
the title compound was obtained as the TFA salt (0.017 g, 21%
1
yield). H NMR (DMSO-d6) δ 12.58 (bs, 1H), 10.18 (bs, 1H),
8.50 (bs, 1H), 8.31 (s, 1H), 7.20 (s, 2H), 4.4 (bs, 2H), 4.0 (bs,
2H), 3.4 (bs, 2H), 3.2 (bs, 2H). HRMS [M + H]+ calcd:
302.1076, found: 302.1103. HPLC (method A): 99% purity.
HPLC (method B): 98.5% purity.
4-({2-[(5-Cyano-1,3-thiazol-2-yl)amino]-4-pyridinyl}-
methyl)-1-piperazinecarboxamide (18). The procedure was
followed as for 15 but employing 1-piperazinecarboxamide
(0.144 g, 1.12 mmol), 1 mL of DMSO, and 2-(4-chlorometh-
ylpyridin-2-ylamino)thiazole-5-carbonitrile (0.070 g, 0.28 mmol).
After preparative reverse phase HPLC, the title compound was
obtained as the TFA salt (0.0249 g, 18% yield). 1H NMR (CD3-
OD) δ 8.49 (d, 1H, J ) 5.0 Hz), 8.06 (s, 1H), 7.15 (m, 2H), 4.25
(s, 2H), 3.64 (bs, 4H), 3.15 (s, 4H). HRMS [M + H]+ calcd:
344.1293, found: 344.1250. HPLC (method A): 97.9% purity.
HPLC (method B): 97.5% purity.
4-({2-[(5-Cyano-1,3-thiazol-2-yl)amino]-4-pyridinyl}-
methyl)-N,N-dimethyl-1-piperazinecarboxamide (19). The
procedure was followed as for 15 but employing N,N-dimethyl-
1-piperazinecarboxamide (0.298 g, 1.90 mmol), 1 mL of DMSO,
and 2-(4-chloromethylpyridin-2-ylamino)thiazole-5-carbonitrile
(0.119 g, 0.47 mmol). The reaction was diluted with water and
the resulting precipitate collected by filtration. The solid was
washed with water and hexanes and then air-dried overnight
to afford the free base (0.0913 g, 49% yield). Free base: 1H
NMR (CD3OD) δ 8.33 (d, 1H, J ) 5.0 Hz), 8.03 (s, 1H), 7.08 (s,
1H), 7.04 (d, 1H, J ) 5.0 Hz), 3.58 (s, 1H), 3.29 (t, 4H, J ) 6.0
Hz), 2.84 (s, 6H), 2.49 (t, 4H, J ) Hz). HRMS [M + H]+ calcd:
372.1607, found: 372.1611. HPLC (method A): 95.2% purity.
HPLC (method B): 98% purity.
4-[2-(5-Cyanothiazol-2-ylamino)pyridin-4-ylmethyl]pip-
erazine-1-carboxylic Acid Methylamide (20). 2-{[4-(Chlo-
romethyl)pyridin-2-yl]amino}-1,3-thiazole-5-carbonitrile (8.00
g, 31.9 mmol) was stirred in 60 mL of DMSO. 1-[(Methylami-
no)carbonyl]piperazin-4-ium chloride (11.5 g, 63.8 mmol) was
added, followed by addition of triethylamine (13.34 mL, 95.7
mmol). The reaction was allowed to stir at room temperature
for 15 h, at which time an additional 2.00 g piperazine
hydrochloride (11.1 mmol) and 6.6 mL of Et3N (48 mmol) were
added. After 1 h the reaction was diluted with 300 mL of water.
The resulting precipitate was filtered, washed with water, and
air-dried. The solid was purified by flash chromatography
(eluted with 92:8 DCM/MeOH), and the resulting still some-
what impure product was purified by reverse phase prepara-
tive HPLC. The resulting TFA salt was dissolved in water,
and the resulting solution was adjusted to pH 7 with saturated
aqueous NaHCO3. The resulting precipitate was filtered and
air-dried to afford 9.40 g of the title compound (82% yield). 1H
NMR (DMSO-d6) δ 12.20 (bs, 1H), 8.32 (d, 1H, J ) 5.49 Hz),
8.26 (s, 1H), 7.13 (s, 1H), 7.03 (d, 1H, J ) 5.19 Hz), 6.42 (bd,
1H, J ) 4.27 Hz), 3.52 (s, 2H), 3.29 (m, 4H), 2.51 (d, 3H, J )
4.27 Hz), 2.33 (m, 4H). [M + H]+ ) 358.1443. Anal. (C16H19N7-
OS) C, H, N.
2-(4-Hydroxymethylpyridin-2-ylamino)thiazole-5-car-
bonitrile. 2-[4-(tert-Butyldimethylsilanyloxymethyl)pyridin-
2-ylamino]thiazole-5-carbonitrile (23.4 g, 67.5 mmol) was
dissolved in 130 mL of anhydrous THF. Hydrogen fluoride-
pyridine (HF ∼70% pyridine ∼30%, 70 mL) was added, and
the reaction was stirred for 1 h. A precipitate formed and was
filtered and washed with THF. The filtrate was diluted with
1 L of water and filtered. The solid was air-dried and combined
with the solid filtered from the reaction to afford 16.0 g (101%)
1
of the title compound. H NMR (DMSO-d6) δ 12.23 (bs, 1H),
8.30 (d, 1H, J ) 5.3 Hz), 8.26 (s, 1H), 7.15 (s, 1H), 6.99 (d, 1H,
J ) 5.3 Hz), 5.49 (t, 1H, J ) 5.7 Hz) 4.54 (d, 2H, J ) 5.7 Hz).
2-(4-Chloromethylpyridin-2-ylamino)thiazole-5-carbo-
nitrile (7, R1 ) H, R2 ) CN). 2-(4-Hydroxymethylpyridin-2-
ylamino)thiazole-5-carbonitrile (16.0 g, 68.8 mmol) was stirred
in anhydrous CH2Cl2 (140 mL) under N2. Dimethylformamide
(5.32 mL, 68.8 mmol) was added followed by the addition of
phosphorus oxychloride (6.41 mL, 68.8 mmol). After 4 h the
reaction was quenched by the addition of 200 mL of saturated
aqueous NaHCO3. A precipitate formed which was filtered and
washed with water to provide 15.5 g (90% yield) of the title
compound. 1H NMR (DMSO-d6) δ 12.35 (bs, 1H), 8.40 (d, 1H,
J ) 5.3 Hz), 8.28 (s, 1H), 7.20 (s, 1H), 7.12 (d, 1H, J ) 5.3
Hz), 4.82 (s, 2H).
2-[4-(4-Acetylpiperazin-1-ylmethyl)pyridin-2-ylamino]-
thiazole-5-carbonitrile (15). 1-Acetylpiperazine (0.460 g,
3.59 mmol) was dissolved in 2 mL of anhydrous DMSO. 2-(4-
Chloromethylpyridin-2-ylamino)thiazole-5-carbonitrile (0.300
g, 1.20 mmol) was added, and the solution was stirred for 30
min. The reaction solution was purified by preparative reverse
phase HPLC. The fractions containing the desired compound
were concentrated to dryness to afford the TFA salt (0.260 g,
47.5% yield). 1H NMR (free base, CDCl3) δ 9.94 (bs, 1H), 8.35
(d, 1H, J ) 5.1 Hz), 7.99 (s, 1H), 7.00 (d, 1H, J ) 5.4 Hz), 6.95
(s, 1), 3.66 (t, 2H, 4.8 Hz), 3.56 (s, 2H), 3.52 (t, 2H, J ) 4.9
Hz), 2.50 (t, 2H, J ) 5.0 Hz), 2.45 (t, 2H, J ) 5.0 hz), 2.11 (s,
3H). dec 241-245 °C. Anal. (C18H19N6O3S) C, H, N.
2-(4-Piperazin-1-ylmethylpyridin-2-ylamino)thiazole-
5-carbonitrile (16). The procedure was followed as for 15 but
employing tert-butyl piperazine-1-carboxylate (0.373 g, 2.00
mmol) in 1 mL of anhydrous DMSO and 2-(4-chloromethylpy-
ridin-2-ylamino)thiazole-5-carbonitrile (0.100 g, 0.401 mmol).
The reaction solution was purified by preparative reverse
phase HPLC. The fractions containing the desired compound
were concentrated to dryness to afford the TFA salt (0.052 g,
25% yield). A pure sample of this product (0.365 g, 0.708 mmol)
was stirred in 15 mL of EtOAc, and the mixture was cooled to
0 °C. Hydrogen chloride (g) was bubbled into the solution for
30 min. The reaction was allowed to warm to ambient
temperature and after 2 h was concentrated to dryness. The
residue was purified by preparative reverse phase HPLC. The
resulting TFA salt was stirred in saturated aqueous NaHCO3,
2-{4-[4-(2-Hydroxyethanoyl)piperazin-1-ylmethyl]py-
ridin-2-ylamino}thiazole-5-carbonitrile (21). The proce-
dure was followed as for 15 but employing 1-glycoloylpipera-
zine hydrochloride (259 mg, 1.44 mmol), 2 mL of DMSO, 2-(4-
chloromethylpyridin-2-ylamino)thiazole-5-carbonitrile (180 mg,
0.72 mmol), and diisopropylethylamine (0.38 mL, 2.15 mmol).
After 3 h the mixture was diluted with H2O and extracted with
EtOAc (3×). The combined organic layers were dried (MgSO4),

6370 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Bilodeau et al.
filtered, and concentrated. Flash column chromatography
(gradient, 5-15% EtOH/EtOAc then 5-10% MeOH/CHCl3)
gave the title compound as a pale yellow solid 86 mg, 33%:
1H NMR (DMSO-d6) δ12.19 (s, 1 H), 8.33 (d, 1 H, J ) 5.1 Hz),
8.32 (s, 1 H), 7.14 (s, 1 H), 7.04 (d, 1 H, J ) 5.2 Hz), 4.54 (t, 1
H, J ) 5.6 Hz), 4.08 (d, 2 H, J ) 5.6 Hz), 3.55 (s, 2 H), 3.49 (s,
2 H), 3.36 (s, 2 H), 2.38 (s, 4 H); HRMS [M + H]+ calcd:
359.1290, found: 359.1285. HPLC (method A): 97.2% purity.
HPLC (method B): 98.3% purity.
stirred overnight. The bulk of the DMF was removed in vacuo,
and the residue was diluted with saturated aqueous NaHCO3.
The mixture was extracted 4 × 100 mL of DCM, and the
combined extracts were dried over Na2SO4, filtered, and
concentrated. The residue was purified by flash column
chromatography (eluting with 93:7 hexanes/EtOAc) to afford
8.89 g of the title compound as a white solid. 1H NMR (CDCl3)
δ 8.23 (d, 1H, J ) 4.9 Hz), 7.40 (d, 1H, J ) 4.9 Hz), 4.69 (s,
2H), 2.27 (s, 3H), 0.96 (s, 9H), 0.13 (s, 6H).
4-(tert-Butyldimethylsilanyloxymethyl)-3-methylpyri-
din-2-ylamine (5, R1 ) Me). An oven-dried flask under N2
was charged with 4-(tert-butyldimethylsilanyloxymethyl)-2-
chloro-3-methylpyridine (8.89 g, 32.7 mmol), NaOtBu (4.40 g,
45.8 mmol), tris(dibenzylideneacetone)dipalladium(0) (0.599 g,
0.65 mmol), racemic-2,2′-bis(diphenylphosphino)-1,1′-binaph-
thyl (1.22 g, 1.96 mmol), and anhydrous toluene (50 mL).
Benzophenone imine (6.58 mL, 39.2 mmol) was added, and
the reaction was heated to 80 °C. After 3 h the reaction was
allowed to cool to room temperature and was diluted with 50
mL of diethyl ether. The resulting mixture was filtered through
Celite, washing with ether. The filtrate was concentrated and
redissolved in 10 mL of MeOH, and hydroxylamine (50%
aqueous, 3.00 mL, 98.1 mmol) was added. After stirring
overnight, an additional 0.680 mL of hydroxylamine solution
was added. After 17 h the solution was concentrated in vacuo.
The residue was purified by flash column chromatography
(hexane/EtOAc) to afford 6.83 g (83% yield) of the title product.
1H NMR (CDCl3) δ 7.80 (d, 1H), 6.85 (d, 1H), 4.65 (s, 2H), 4.35
(bs, 2H), 2.00 (s, 3H), 0.95 (s, 9H), 0.15 (s, 6H).
2-[(4-{[4-(Methylsulfonyl)piperidin-1-yl]methyl}pyridin-
2-yl)amino]-1,3-thiazole-5-carbonitrile (22). 2-{[4-(Chlo-
romethyl)pyridin-2-yl]amino}-1,3-thiazole-5-carbonitrile (3.490
g, 13.92 mmol) was dissolved in 20 mL of DMSO. 4-(Methyl-
sulfonyl)piperidine (3.409 g, 20.88 mmol) and N-ethyl-N,N-
diisopropylamine (9.70 mL, 55.68 mmol) were added, and the
solution was stirred for 17 h. The reaction was diluted with
water, and the resulting precipitate was filtered and washed
with water. The precipitate was then purified by flash column
chromatography (gradient, 3-10% MeOH/CH2Cl2). The frac-
tions containing the desired compound were concentrated to
dryness to afford 1.82 g of the title compound (35% yield). 1H
NMR (DMSO-d6) δ 12.18 (bs, 1H), 8.32 (d, 1H, J ) 5.19 Hz),
8.26 (s, 1H), 7.15 (s, 1H), 7.01 (d, 1H, J ) 5.19 Hz), 3.54 (s,
2H), 3.14-3.09 (m, 1H), 2.93 (s overlapping m, 5H), 2.08-1.98
(m, 4H), 1.68-1.55 (m, 2H). HRMS [M + H]+ calcd: 378.1058,
found: 378.1057. HPLC (method A): 100% purity. HPLC
(method B): 100% purity.
3-Methyl-1-oxyisonicotinic Acid Ethyl Ester. 3-Meth-
ylisonicotinic acid ethyl ester18 (11.2 g, 67.8 mmol) was
dissolved in 24 mL of dichloromethane. Methyl trioxorhenium
(84.5 mg, 0.34 mmol) was added followed by the addition of
hydrogen peroxide (30% aqueous, 13.8 mL, 135 mmol). The
reaction was allowed to stir overnight. Additional methyl
trioxorhenium (84.5 mg, 0.34 mmol) and hydrogen peroxide
(30% aqueous, 13.8 mL, 135 mmol) were added. After 7 h, 25
mg of MnO2 was added to the reaction (vigorous bubbling
occurred). After bubbling had subsided (30 min), the reaction
was diluted with water and extracted 3× with dichlo-
romethane. The combined extracts were dried over Na2SO4,
filtered, and concentrated to afford 11.1 g (90% yield) of the
title compound as a white solid. 1H NMR (CDCl3) δ 8.06 (s
overlapping with d 2H), 7.82 (d, 1H), 4.37 (q, 2H, J ) 7.0 Hz),
2.55 (s, 3H), 1.40 (t, 3H, J ) 7.0 Hz).
2-[4-(tert-Butyldimethylsilanyloxymethyl)-3-methylpy-
ridin-2-ylamino]thiazole-5-carbonitrile (6, R1 ) Me, R2
) CN). An oven-dried flask under N2 was charged with NaH
(60% dispersion, 2.60 g, 108 mmol). Anhydrous THF (50 mL)
was added followed by 4-(tert-butyldimethylsilanyloxymethyl)-
3-methylpyridin-2-ylamine (6.83 g, 27.1 mmol) and 2-chloro-
5-cyanothiazole (4.69 g, 32.5 mmol). The resulting solution was
heated to reflux. After 3 h the reaction was allowed to cool to
room temperature and diluted with water. The pH was
adjusted to 7 with 1 M HCl. The resulting precipitate was
filtered and washed with water to afford 10.5 g (107% yield)
1
of the title compound. H NMR (CDCl3) δ 8.41 (bs, 1H), 8.27
(d, 1H, J ) 5.2 Hz), 7.94 (s, 1H), 7.19 (d, 1H, J ) 5.2 Hz), 4.74
(s, 2H), 2.21 (s, 3H), 0.96 (s, 9H), 0.13 (s, 6H).
2-(4-Hydroxymethyl-3-methylpyridin-2-ylamino)thia-
zole-5-carbonitrile. 2-[4-(tert-Butyldimethylsilanyloxymethyl)-
3-methylpyridin-2-ylamino]thiazole-5-carbonitrile (10.5 g, 29.1
mmol) was dissolved in 60 mL of THF. HF-pyridine (∼70%
HF, ∼30% pyridine, 29 mL) was added, and the reaction was
stirred at room temperature. After 4 h the resulting precipitate
was filtered and washed with THF to afford 4.39 g (61% yield)
of the pure title compound. 1H NMR (DMSO-d6) δ 11.40 (s,
1H), 8.30 (s, 1H), 8.26 (d, 1H, J ) 5.5 Hz), 7.21 (d, 1H, J ) 5.2
Hz), 4.56 (s, 2H), 2.24 (s, 3H). MS [M + H]+ ) 247.1.
2-(4-Chloromethyl-3-methylpyridin-2-ylamino)thiazole-
5-carbonitrile (7, R1 ) Me, R2 ) CN). 2-(4-Hydroxymethyl-
3-methylpyridin-2-ylamino)thiazole-5-carbonitrile (4.39 g, 17.8
mmol) was stirred in 40 mL of anhydrous DCM. Anhydrous
N,N-dimethylformamide (1.38 mL, 17.8 mmol) was added
followed by phosphorus oxychloride (1.66 mL, 17.8 mmol).
After 2 h the reaction was concentrated in vacuo and quenched
with saturated NaHCO3 (aq). The resulting orange solid was
filtered, washed with water, and dried to afford 3.98 g (84%
yield) of the title compound. 1H NMR (DMSO-d6) δ 11.54 (bs,
1H), 8.33 (s, 1H), 8.29 (d, 1H, J ) 5.1 Hz), 7.19 (d, 1H, J ) 5.1
Hz), 4.84 (s, 2H), 2.39 (s, 3H).
4-[2-(5-Cyanothiazol-2-ylamino)-3-methylpyridin-4-yl-
methyl]piperazine-1-carboxylic Acid Methylamide (23).
2-(4-Chloromethyl-3-methylpyridin-2-ylamino)thiazole-5-car-
bonitrile (0.328 g, 1.24 mmol) and 4-methylcarbamoyl-piper-
azin-1-ium chloride (0.445 g, 2.48 mmol) were dissolved in 3
mL of DMSO. Diisopropylethylamine (0.863 mL, 4.96 mmol)
was added, and the reaction was stirred for 5 h. The reaction
mixture was then directly loaded onto a reverse phase puri-
fication system, which afforded the title compound as the TFA
(2-Chloro-3-methylpyridin-4-yl)methanol. 3-Methyl-1-
oxyisonicotinic acid ethyl ester (10.0 g, 55.3 mmol) was stirred
in POCl3 (25.8 mL, 42.4 g, 277 mmol) in a flask equipped with
a reflux condenser and a drying tube. The resulting mixture
was heated to reflux. After 3 h the reaction was cooled to room
temperature. The excess POCl3 was removed in vacuo. The
residue was diluted with dichloromethane and washed with
aqueous NaHCO3 (sat). The aqueous phase was extracted 2×
with dichloromethane, dried over Na2SO4, filtered, and con-
centrated. This afforded two isomeric products as a 2.5:1
mixture as a tan solid (9.76 g, 88% yield). A portion of this
mixture (4.81 g, 24.1 mmol) was dissolved in 70 mL of
anhydrous THF in an oven-dried flask under N2. A reflux
condenser was added, and a solution of LiBH4 in THF (2 M,
14.5 mL, 28.9 mmol) was added. The reaction was heated to
reflux for 2 h, allowed to cooled to room temperature, and was
then quenched by the addition of 1 M HCl (aq). The solution
was extracted 3× with dichloromethane. The combined ex-
tracts were dried over Na2SO4, filtered and concentrated. The
resulting residue was purified by flash column chromatogra-
phy (dissolved sample in DCM, eluted with 4:1 DCM/EtOAc)
which afforded 0.77 g (20% yield) of the undesired (2-chloro-
5-methylpyridin-4-yl)-methanol and 1.82 g (48% yield) of the
1
desired title compound. H NMR (CDCl3) δ 8.22 (d, 1H, J )
4.9 Hz), 7.44 (d, 1H, J ) 4.9 Hz), 5.53 (t, 1H, J ) 5.4 Hz), 4.56
(d, 2H, J ) 6.1 Hz), 2.23 (s, 3H).
4-(tert-Butyldimethylsilanyloxymethyl)-2-chloro-3-me-
thylpyridine (11). (2-Chloro-3-methylpyridin-4-yl)-methanol
(6.71 g, 42.6 mmol), tert-butyldimethylsilyl chloride (7.06 g,
46.8 mmol), and imidazole (3.48 g, 51.1) were dissolved in 80
mL of anhydrous DMF in an oven-dried flask under N2 and

VEGF Receptor Tyrosine Kinase Inhibitors
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25 6371
salt (0.345 g, 57% yield). 1H NMR (DMSO-d6) δ 11.60 (s, 1H),
9.77 (bs, 1H), 8.36 (s overlapping with d, 2H), 7.23 (d, 1H, J )
5.2 Hz), 6.69 (s, 1H), 4.44 (s, 2H), 4.03 (s, 2H), 3.36 (s, 2H),
3.06 (s, 4H), 2.59 (s, 3H), 2.42 (s, 3H). HRMS [M + H]+ calcd:
372.1606, found: 372.1597. HPLC (method A): 100%, HPLC
(method B): 100% purity.
48 h. In a preliminary assessment of the excretion and
metabolic pathways, bile and urine were collected from one
dogs for 72 and 120 h, respectively, after iv administration of
a 2.4-mg/kg dose (solution in DMSO) of [14C]20. In a prelimi-
nary assessment of the routes of excretion and metabolic
pathways of 20, urine and feces were collected from one
monkey for 96 h and bile for 72 h after iv administration of a
1.6-mg/kg dose (solution in DMSO) of [14C]20.
The effect of 20 and 23 on the metabolism of probe
substrates for CYP 1A2 (substrate: phenacetin), 2C8 (pacli-
taxel), 2C9 (diclofenac), 2C19 (S-mephenytoin), 2D6 (bu-
furalol), and 3A4 (testosterone) was measured in human liver
microsomes.
N-(5-Cyano-1,3-thiazol-2-yl)-N-[4-({4-[(methylamino)-
carbonyl]piperazin-1-yl}methyl)pyridin-2-yl]-D-glucopy-
ranuronosylamine (24). NMR data for the N-glucuronide
conjugate was acquired on a 500 MHz Inova spectrometer
(Varian Inc. Palo Alto, CA) equipped with a 3-mm MIDG probe
(Varian Inc.) at 25 °C. The N-glucuronide conjugate, isolated
from Rhesus urine, was dissolved in 160 µL of CD3OD and
transferred to a 3-mm NMR tube. The site of glucuronide
1
conjugation was inferred from the 1D H, TOCSY, gHMQC,
Acknowledgment. We thank Dr. Conrad Raab for
the preparation of [14C]20 and Matthew M. Zrada and
Kenneth D. Anderson for log P and protein binding
determinations.
gHMBC data sets. All chemical shifts are referenced to the
CD2HOD resonance set at δH 3.33 ppm/δC 48.0 ppm.
NMR data on the N-glucuronide conjugate showed the
presence of all peaks expected from 20 and also those from
the glucuronide moiety. The glucuronide anomeric CH (2′′)
appeared at δH 6.41 ppm (d, 7.4 Hz) and at δC 73.5 ppm.
Pyridine C-2′ and cyanothiazole C-2 appeared at δC 137.6 and
153.6 ppm, respectively. gHMBC correlations were observed
from H2′′ to both C-2′ and C-2. gHMBC correlations were also
observed from pyridine H-6′ (8.48 ppm) to C-2′ and from
cyanothiazole H-1 (8.06 ppm) to C-2. The combined NMR data
placed the N-glucuronide at the bridge nitrogen connecting the
pyridine and the cyanothiazole rings.
Supporting Information Available: Table of elemental
analysis data. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) (a) Thomas, K. A. Vascular endothelial growth factor, a potent
and selective angiogenic agent. J. Biol. Chem. 1996, 271, 603-
606. (b) Ferrara, N.; Davis-Smyth, T. The biology of vascular
endothelial growth factor. Endocrine Rev. 1997, 18, 4-25.
(2) Fox, S. B.; Gasparini, G.; Harris, A. L. Angiogenesis: pathologi-
cal, prognostic, and growth factor pathways and their link to
trial design and anticancer drugs. Lancet Oncol. 2001, 2, 278-
289.
N-glucuronide conjugate NMR data, listed as position, δH
(ppm), δC (ppm). Thiazole: C-2, 153.6; CH-1, 8.06, 133.4; C-5,
84.2; 5-CN, 99.3. Pyridine: C-2′, 137.6; CH-3′, 7.89, 100.9; C-4′,
136.5; CH-5′, 7.27, 105.9; CH-6′, 8.48, 130.9. Piperazine: CH2,
3.49, 28.2; CH2, 2.73, 11.4. Methylamide: CdO, 145.6; N-Me,
2.73, 11.4. Glucuronide: CH-1′′, 6.41, 73.5; CH-2′′, 4.02, 54.1;
CH-3′′, 3.69, 56.9; CH-4′′, 3.59, 62.5; CH-5′′, 4.08, 63.1.
Biological Methods. Enzyme, cell-based, and in vivo
kinase assays were performed according to previously de-
scribed procedures.7
The affinity of compounds for the IKr channel was evaluated
in radioligand competition experiments analogous to [3H]-
dofetilide binding assays.19 The human ERG potassium chan-
nel was stably expressed in 293 embryonic kidney cells,20 and
plasma membrane fractions prepared from these cells were
used for competition experiments with [35S]-MK499.21 Results
are reported as inflection points (IP) for the test compounds
to inhibit binding of the radioligand. Most determinations were
carried out at least twice, and the repeats were generally
within 2-fold of the original value. Standard deviations from
the mean value are shown for IP values determined three or
more times.
(3) Neufeld, G.; Cohen, T.; Gengrinovitch, S.; Poltorak, Z. Vascular
endothelial growth factor (VEGF) and its receptors. FASEB
1999, 13, 9-22.
(4) (a) Yang, J. C.; Haworth, L.; Sherry, R. M.; Hwu, P.; Schwartzen-
truber, D. J.; Topalian, S. L.; Steinberg, S. M.; Chen, H. X.;
Rosenberg, S. A. A randomized trial of bevacizumab, an anti-
vascular endothelial growth factor antibody, for metastatic renal
cancer. N. Engl. J. Med. 2003, 349, 427-34. (b) McCarty M.
Antiangiogenesis drug promising for metastatic colorectal can-
cer: New treatments for colorectal cancer might improve
patients’ survival, investigators reported at ASCO. Lancet 2003,
361, 1959.
(5) For recent reviews, see: (a) Bilodeau, M. T.; Fraley, M. E.;
Hartman, G. D. Kinaseinsetdomain-containing receptor kinase
inhibitors as antiangiogenic agents. Expert Opin. Investig. Drugs
2002, 11, 737-745. (b) Sepp-Lorenzino, L.; Thomas, K. A.
Antiangiogenic agents targeting vascular endothelial growth
factor and its receptors in clinical development. Expert Opin.
Investig. Drugs 2002, 11, 1447-1465.
(6) Bilodeau, M. T.; Rodman, L. D.; McGaughey, G. B.; Coll, K. E.;
Koester, T. J.; Hoffman, W. F.; Hungate, R. W.; Kendall, R. L.;
McFall, R. C.; Rickert, K. W.; Rutledge, R. Z.; Thomas, K. A.
The Discovery of N-(1,3-Thiazol-2-yl)pyridin-2-amines as Potent
Inhibitors of KDR Kinase. Bioorg. Med. Chem. Lett. 2004, 14,
2941-2945.
(7) Rickert, K.; Coll, K. E.; Mao, X.; McFall, R. C.; Pan, B.-S.; Zeng,
Q.; Johnson, A.; Sepp-Lorenzino, L.; Shipman, J. M.; Mc-
Gaughey, G. B.; Kendall, R. L.; Huckle, W. R.; Brunner, J. E.;
Anderson, K. D.; Fraley, M. E.; Hoffman, W. F.; Bilodeau, M.
T.; Hartman, G. D.; Heimbrook, D. C.; Gibbs, J. B.; Kohl, N.;
Thomas, K. A. Activities of a potent inhibitor of the unactivated
form of the human vascular endothelial growth factor receptor
KDR. J. Biol. Chem., submitted.
(8) Roden, D. M. Drug-induced prolongation of the QT interval. N.
Engl. J. Med. 2004, 350, 1013-1022.
(9) Cavero, I.; Mestre, M.; Guillon, J.-M.; Crumb, W. Drugs that
prolong QT interval as an unwanted side effect: assessing their
likelihood of inducing hazardous cardiac dysrhythmias. Exp.
Opin. Pharmacother. 2000, 1, 947-973.
(10) De Ponti, F.; Poluzzi, E.; Montanaor, N. Organising evidence
on QT prolongation and occurrence of Torsades de Pointes with
nonantiarrythmic drugs: a call for consensus. Eur. J. Clin.
Pharmacol. 2001, 57, 185-209.
(11) McElhinney, R. S.; Donnelly, D. J.; McCormick, J. E.; Kelly, J.;
Watson, A. J.; Rafferty, J. A.; Elder, R. H.; Middleton, M. R.;
Willington, M. A.; McMurry, T. B. H. Margison, G. P. Inactiva-
tion of O6-alkylguanine-DNA alkyltransferase. 1. Novel O6-
(hetarylmethyl)guanines having basic rings in the side chain J.
Med. Chem. 1998, 41, 5265-5271.
Class III electrophysiologic activity was assessed in anes-
thetized dogs with the compound of interest being adminis-
tered by intravenous infusion in an escalating-dose protocol.
Plasma levels of compound and electrocardiogram data were
determined before and during the infusions. The effects on the
QTc interval (the duration from the start of the QRS complex
to the end of the T wave, corrected for heart rate) were
correlated with drug concentrations in plasma.
For rat pharmacokinetics, fasted male Sprague-Dawley
rats were given a 10 mg/kg dose by oral gavage and a 2 mg/kg
dose by bolus intravenous injection. Oral bioavailability was
determined using a solution of the mesylate salt in 0.05 M
citric acid and the iv dose was administered as a DMSO
solution. The bioavailability was assessed in fasted beagle dogs
as a solution in 0.05 M citric acid or HCl-acidified methocel.
For iv study a 1-mg/kg dose was administered intravenously
as a DMSO solution. The bioavailability of a 1 mg/kg oral dose
was assessed in fasted Rhesus monkeys using a solution in
0.05 M citric acid. In a crossover design, a 1 mg/kg dose was
administered intravenously as a DMSO solution on a separate
occasion.
The disposition of [14C]20 was assessed in bile-duct cannu-
lated rats. Following iv administration of a 4-mg/kg dose
(solution in DMSO), bile and urine were collected over ice for

6372 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Bilodeau et al.
(12) Wolfe, J. P.; Ahman, J.; Sadighi, J. P.; Singer, R. A.; Buchwald,
S. L. An ammonia equivalent for the palladium-catalyzed
amination of aryl halides and triflates. Tetrahedron Lett. 1997,
38, 6367-6370.
(13) Coperet, C.; Adolfsson, H.; Khuong, T.-A. V.; Yudin, A. K.;
Sharpless, K. B. A simple and efficient method for the prepara-
tion of pyridine N-oxides. J. Org. Chem. 1998, 63, 1740-1741.
(14) Netzer, R.; Ebneth, A.; Bischoff, U.; pongs, O. Screening lead
compounds for QT interval prolongation. Drug Discovery Today
2001, 6, 78-84.
(15) Skobe, M., Hawighorst, T., Jackson, D. G., Prevo, R., Janes, L.,
Velasco, P., Riccardi, L., Alitalo, K., Claffey, K., and Detmar,
M. Induction of tumor lymphangiogenesis by VEGF-C promotes
breast cancer metastasis. Nat. Med. 2001, 7, 192-198.
(16) Hafez, E. A. A.; Abed, N. M.; Elsakka, I. A. Reactions with
heterocyclic diazonium salts. Synthesis of several new azolyl-
hydrazones. J. Heterocycl.Chem. 1983, 20, 285-288.
(17) Borthwick, A. D.; Foxton, M. W.; Gray, B. V.; Gregory, G. I.;
Seale, P. W.; Warburton, W. K. Some 2-nitrothiazole. J. Chem.
Soc., Perkin Trans. 1 1973, 2769-2772.
(18) Clarke, K.; Goulding, J.; Scrowston, R. M. Preparation of some
thiopyranopyridine derivatives. J. Chem. Soc., Perkin Trans. 1
1984, 1501-1505.
(19) Fiset, C.; Feng, Z.-P.; Wang, L.; Sheldon, R. S.; Duff, H. J. [3H]-
Dofetilide Binding: Biological Models that Manifest Solely the
High or the Low Affinity Binding Site. J. Mol. Cell. Cardiol.
1996, 28, 1085-1096.
(20) Zhou, Z.; Gong, Q.; Ye, B.; Fan, Z.; Makielski, J. C.; Robertson,
G. A.; January, C. T. Properties of HERG Channels Stably
Expressed in HEK 293 Cells Studied at Physiological Temper-
ature. Biophys. J. 1998, 74, 230-241.
(21) Claremon, D. A.; Baldwin, J. J.; Elliott, J. M.; Remy, D. C.;
Ponticello, G. S.; Selnick, H. G.; Lynch, J. J., Jr.; Sanguinetti,
M. C.; Selective IKr Potassium Channel Blockers as Class III
Antiarrhythmic Agents. In Perspectives in Medicinal Chemistry;
Testa, B., Kyburz, E., Fuhrer, W., Giger, R., Eds.; Verlag
Helvetica Chimica Acta: Basel, 1993; pp 389-404.
JM049697F